If a production computer system crashes, the system operator would like to recover the production system and get it back up and running as quickly as possible. However, many such production systems contain a vast amount of data, so the recovery process consequently could take several hours or more. This is simply not acceptable in many production environments, and an immediate recovery is desired by most systems operators and users.